Master Informatique 1 Semantic Web Technologies Part 0Course Organization Semantic Web Technologies Werner Nutt

Master Informatique 2 Semantic Web Technologies Part 0Course Organization The Semantic Web Idea Semantic Technologies

Master Informatique 3 Semantic Web Technologies Part 0Course Organization The Semantic Web Idea Semantic Technologies

Master Informatique 4 Semantic Web Technologies Part 0Course Organization The Web All aspects of daily life are being “webized”: everyday information (news, weather forecast) entertainment buying and selling administration (eGovernment) education (eLearning, eEducation) social contacts (social networking plattforms, dating sites)

Master Informatique 5 Semantic Web Technologies Part 0Course Organization Why Semantic Web? Syntax vs. Semantik –“4+) = (“ syntactically wrong, semantically … –“3 + 4 = 12” syntactically correct, semantically wrong –“3 + 4 = 7” syntactically correct, semantically correct (from greek συνταξις composition,sentential structure) denotes the (normative) structure of data, i.e., it characterizes what makes data “well-formed” (greek σημαυτικoς belonging to the sign) denotes the meaning of data, i.e., it characterizes what conclusions can be drawn from it. Syntax Semantik

Master Informatique 6 Semantic Web Technologies Part 0Course Organization Web of Documents Primary objects: documents Links between documents Degree of structure in data: low Semantics of content: implicit Designed for: human consumption

Master Informatique 7 Semantic Web Technologies Part 0Course Organization Problems of the Web wealth of information … targeted at human users

Master Informatique 8 Semantic Web Technologies Part 0Course Organization Problems of the Web wealth of information … targeted at human users

Master Informatique 9 Semantic Web Technologies Part 0Course Organization Pages with Computer-readable Layout Information Sebastian Rudolph Position Head of the Group address Telephone +49 (0) Office Room 2035, Nöthnitzer Straße 46, Dresden-Räcknitz External page rudolph.de/ Curriculum vitae

Master Informatique 10 Semantic Web Technologies Part 0Course Organization Web of documents

Master Informatique 11 Semantic Web Technologies Part 0Course Organization Problems of the Web localizing information problematic today’s search engines are good, but still keyword-based

Master Informatique 12 Semantic Web Technologies Part 0Course Organization Problems of the Web How about this query: –How many romantic comedy Hollywood movies are directed by a person who is born in a city that has average temperature above 15 degrees!? You need to: –Find reliable sources containing facts about movies (genre & director), birthplaces of famous artists/directors, average temperature of cities across the world, etc.  The result: several lists of thousands of facts –Integrate all the data, join the facts that come from heterogeneous sources Even if possible, it may take hours to answer just a single query!

Master Informatique 13 Semantic Web Technologies Part 0Course Organization Problems of the Web Heterogeneity of present information on diverse levels: character encoding (e.g.ASCII vs.Unicode) used natural languages positioning of information on web pages desirable: information integration across the Web

Master Informatique 14 Semantic Web Technologies Part 0Course Organization Problems of the Web implicit knowledge, i.e. many pieces of information are not provided explicitly, but follow from the combination of the given data requires “logical thinking”  automated deduction

Master Informatique 15 Semantic Web Technologies Part 0Course Organization Problems of the Web Approaches toward a solution: ① Ad hoc: Deployment of AI methods (most notably NLP techniques) to evaluate existing unstructured information on the Web ① A priori: structure information on the Web at authoring time in a way facilitating later automated deployment  Semantic Web

Master Informatique 16 Semantic Web Technologies Part 0Course Organization + The Vision I have a dream for the Web in which computers become capable of analyzing all the data on the Web - the content, links, and transactions between people and computers. A Semantic Web, which should make this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to ma- chines. The intelligent agents people have touted for ages will finally materialize. Barners-Lee, 1999

Master Informatique 17 Semantic Web Technologies Part 0Course Organization Semantic Web Two essential prerequisites for the implementation: ① open standards for describing information clearly defined flexible extendable ② methods for retrieving information from such descriptions

Master Informatique 18 Semantic Web Technologies Part 0Course Organization Semantic Web: Standards 1994 First public presentation of the Semantic Web idea 1998 Start of standardization of data model (RDF) and a first ontology languages (RDFS) at W3C 2000 Start of large research projects about ontologies in the US and Europe 2002 Start of standardization of a new ontology language (OWL) based on research results 2004 Finalization of the standard for data (RDF) and ontology (OWL) 2008 Standardization of a query language (SPARQL) 2009 Extension of OWL to OWL Standard Rule Interchange Format (RIF) 2013 SPARQL 1.1

Master Informatique 19 Semantic Web Technologies Part 0Course Organization The Semantic Web Idea Semantic Technologies

Master Informatique 20 Semantic Web Technologies Part 0Course Organization Not only about the Web The Semantic Web vision has generated technologies that are applied outside the web context including: –Enterprise intelligence –Government –Research (Bio, Geo, Cultural heritage, etc.) –Software development –… Semantic technologies provide flexible and powerful tools to accomplish things that were not possible or not practical in the past.

Master Informatique 21 Semantic Web Technologies Part 0Course Organization Introduction to the Semantic Web Approach 21 How does the Semantic Web approach help us merge data sets infer new relations, and integrate outside data sources? How does the Semantic Web approach help us merge data sets infer new relations, and integrate outside data sources?

Master Informatique 22 Semantic Web Technologies Part 0Course Organization Sketch of Data Integration with SWT 1.Map the various data onto an abstract data representation  Make the data independent of its internal representation … 2.Merge the resulting representations 3.Start querying the whole merged set  We can answer queries not possible on the individual data sets 22

Master Informatique 23 Semantic Web Technologies Part 0Course Organization Data set “A”: A simple Book Store 23 Books Authors Publishers

Master Informatique 24 Semantic Web Technologies Part 0Course Organization 1 st : Export your data as a labeled graph 24

Master Informatique 25 Semantic Web Technologies Part 0Course Organization Remarks on the Data Export  Data export does not necessarily mean physical conversion of the data  Relations can be virtual, generated on-the-fly at query time  via SQL “bridges”  scraping HTML pages  extracting data from Excel sheets  etc.  One can export part of the data 25

Master Informatique 26 Semantic Web Technologies Part 0Course Organization Data set “F”: Another Book Store’s Data 26

Master Informatique 27 Semantic Web Technologies Part 0Course Organization 2 nd : Export your second set of data 27

Master Informatique 28 Semantic Web Technologies Part 0Course Organization 3 rd : Start merging your data 28

Master Informatique 29 Semantic Web Technologies Part 0Course Organization 3 rd : start merging your data (cont’d) 29

Master Informatique 30 Semantic Web Technologies Part 0Course Organization 4 th : Merge identical resources 30

Master Informatique 31 Semantic Web Technologies Part 0Course Organization Start posing queries… User of data set “F” can now ask queries like: –“What is the title of the original version of Le Palais des miroirs?” This information is not in the data set “F”... …but can be retrieved after merging with data set “A”! 31

Master Informatique 32 Semantic Web Technologies Part 0Course Organization 5 th : Query the merged data set 32

Master Informatique 33 Semantic Web Technologies Part 0Course Organization However, we can achieve more … We “know” that a:author and f:auteur are really the same But our automatic merge does not know that! Let us add some extra information to the merged data: –a:author is equivalent to f:auteur –both denote a Person, a category (type) for certain resources –a:name and f:nom are equivalent to foaf:name 33

Master Informatique 34 Semantic Web Technologies Part 0Course Organization 3 rd revisited: Use the extra knowledge 34

Master Informatique 35 Semantic Web Technologies Part 0Course Organization Start making richer queries! User of data set “F” can now query: –“What is the home page of the ‘auteur’ of the orignal version of Le Palais des miroirs ?” The information is not in data set “F” or “A”… …but was made available by: –Merging data sets “A” and “F” –Adding three simple “glue” statements 35

Master Informatique 36 Semantic Web Technologies Part 0Course Organization 6 th : Richer queries 36

Master Informatique 37 Semantic Web Technologies Part 0Course Organization Bring in other data sources We can integrate new information into our merged data set from other sources –e.g. additional information about author Amitav Ghosh Perhaps the largest public source of general knowledge is Wikipedia –Structured data have been extracted from Wikipedia and put into a resource called DBpedia May 12,

Master Informatique 38 Semantic Web Technologies Part 0Course Organization 7 th : Merge with DBpedia data 38 owl:sameAs

Master Informatique 39 Semantic Web Technologies Part 0Course Organization 7 th (cont’d): Merge with DBpedia data 39 owl:sameAs

Master Informatique 40 Semantic Web Technologies Part 0Course Organization 7 th (cont’d): Merge with DBpedia data 40 owl:sameAs

Master Informatique 41 Semantic Web Technologies Part 0Course Organization Rigour Makes the Difference What happened via automatic means is done every day by Web users! The difference: a bit of extra rigour so that machines can do this, too 41

Master Informatique 42 Semantic Web Technologies Part 0Course Organization What Did We Do? We combined different data sets that –may be internal or somewhere on the Web –are of different formats (RDBMS, Excel spreadsheet, (X)HTML, etc) –have different names for the same relations We could combine the data because some URIs were identical –i.e. the ISBNs in this case We could add some simple additional information (the “glue”) to help further merge data sets  We could answer queries that previously could not even be asked 42

Master Informatique 43 Semantic Web Technologies Part 0Course Organization What Did We Do? (cont’d) 43

Master Informatique 44 Semantic Web Technologies Part 0Course Organization The Abstraction Pays off Because… …the graph representation is independent of the details of the native structures …a change in local database schemas, HTML structures, etc. does not affect the whole –“schema independence” …new data, new connections can be added seamlessly & incrementally … it doesn’t matter if you are at the enterprise level or at the web level 44

Master Informatique 45 Semantic Web Technologies Part 0Course Organization So where is the Semantic Web? 45 Semantic Web technologies make such integration possible

Master Informatique 46 Semantic Web Technologies Part 0Course Organization Semantic Technologies in Business SNOMED Oracle (Server) IBM (DB2, Watson) Schema.org Good-relations SAP BBC Best Buy Startups

Master Informatique 47 Semantic Web Technologies Part 0Course Organization Semantic Technologies: Overview A set of technologies and frameworks that enable semantic data management, data integration and the web of data –Resource Description Framework (RDF) –A variety of data interchange formats (e.g., RDF/XML, N3, Turtle, N-Triples) –Semantic languages such as RDF Schema (RDFS), the Web Ontology Language (OWL) and the Query language (SPARQL) –Software infrastructure (RDF/SPARQL frameworks, Triple stores, Data integrators, Query engines, Reasoners) –Publicly available connected dataset and open data initiatives (LOD)

Master Informatique 48 Semantic Web Technologies Part 0Course Organization Reading Material PSW Chapter 1 SWP Part I, Chapter 1 FTW Section 1.4